Research On Task Offloading And Resource Allocation In Mobile Edge Computing

With the rapid development of emerging mobile applications such as augmented reality and virtual reality,it is increasingly difficult for terminal devices to meet the needs of emerging mobile applications for low latency and low energy consumption.In order to solve the above problems,the industry proposes mobile edge computing(MEC).Mobile edge computing technology provides computing power for users at the edge through wireless network.Tasks can be executed locally or unloaded to the mobile edge server,which effectively reduces the delay and energy consumption of terminal devices and improves the quality of user experience.Task unloading and resource allocation are the key research issues in mobile edge computing.However,MEC technology still faces many problems in computing offload.This paper focuses on the research of computing offload and resource allocation in the multi-user scenario of MEC system.This paper analyzes the situation of single task and multi task respectively,and proposes an expected time delay and limited energy algorithm(ETD-LE).The mode factor is set according to whether the user equipment has the mode adjustment function,and the weight coefficient is allocated more reasonably according to the characteristics of terminal equipment and task.In this paper,the task unloading decision,channel allocation,local resource and MEC resource allocation are considered comprehensively,and the problem of minimizing task execution cost is established.This paper first transforms the problem into three sub problems: resource allocation,channel allocation and task unloading decision.By solving the optimal allocation resources of local equipment,the unloading decision is determined according to the cost,and the resource allocation at MEC end is solved by convex optimization.On the basis of satisfying the delay,the optimal channel allocation is obtained by iterative optimization,and then the problem is solved.The experiment shows that compared with the task executed locally or all in the cloud and random strategy,the algorithm can save more time delay and energy consumption cost,and can better meet the user experience.In this paper,a task unloading and resource allocation algorithm(CGA)based on improved heuristic algorithm is proposed for multi MEC nodes in the edge cloud collaborative scenario.Firstly,the algorithm determines whether the task is processed at MEC or cloud based on whether the delay is satisfied,and the node allocation is calculated according to the improved genetic algorithm.The algorithm quantifies the completion time and penalty factors as fitness function.By improving roulette selection strategy,optimizing mutation and crossover operator,the algorithm introduces the pre disaster change strategy,expands the search range,effectively solves the problem that traditional heuristic algorithm is prone to local optimization and is easy to mature early,and significantly shortens the completion time of the task.The simulation results show that compared with the improved differential evolution and the traditional genetic algorithm,the algorithm can save more time,and can alleviate the problem of premature algorithm and weak global search ability.